Epidemiology and Molecular Characterizations of Azole Resistance in Clinical and Environmental Aspergillus fumigatus Isolates from China

Abstract

Azole resistance in Aspergillus fumigatus has emerged as a worldwide public health problem. We sought here to demonstrate the occurrence and characteristics of azole resistance in A. fumigatus from different parts of China. A total of 317 clinical and 144 environmental A. fumigatus isolates from 12 provinces were collected and subjected to screening for azole resistance. Antifungal susceptibility, cyp51A gene sequencing, and genotyping were carried out for all suspected azole-resistant isolates and a subset of azole-susceptible isolates. As a result, 8 (2.5%) clinical and 2 (1.4%) environmental A. fumigatus isolates were identified as azole resistant. Five azole-resistant strains exhibit the TR₃₄/L98H mutation, whereas four carry the TR₃₄/L98H/S297T/F495I mutation in the cyp51A gene. Genetic typing and phylogenetic analysis showed that there was a worldwide clonal expansion of the TR₃₄/L98H isolates, while the TR₃₄/L98H/S297T/F495I isolates from China harbored a distinct genetic background with resistant isolates from other countries. High polymorphisms existed in the cyp51A gene that produced amino acid changes among azole-susceptible A. fumigatus isolates, with N248K being the most common mutation. These data suggest that the wide distribution of azole-resistant A. fumigatus might be attributed to the environmental resistance mechanisms in China.

INTRODUCTION

Aspergillus fumigatus is an important fungal pathogen that causes allergic, chronic, and acute invasive diseases in humans and animals (1). The fungus is ubiquitous in the natural environment and is mainly spread by abundantly produced asexual spores. Azoles are the first-line drugs used in the management of aspergillus diseases (2). However, the clinical use of azoles is threatened by the emergence of azole-resistant A. fumigatus at a global scale in recent years (3). The most common resistance mechanism is characterized by combinations of a tandem repeat (TR₃₄) in the cyp51A promoter and a concomitant mutation in the cyp51A gene itself (L98H), which is believed to be primarily driven by the use of azole fungicides in the environment (4, 5). A new environmental cyp51A-mediated resistance mechanism (TR₄₆/Y121F/T289A), which might confer high level resistance to voriconazole (VRC), has been reported recently in countries from many continents (6–11).

To allow effective measures to be implemented aimed at retaining the use of azoles in clinical medicine and agriculture, more efforts are required to understand how resistance develops in different parts of the world. The emergence of A. fumigatus harboring TR₃₄/L98H/S297T/F495I mutation in China was first reported from a global surveillance study conducted in 2008 and 2009; all eight of the resistant isolates originated from different centers in Hangzhou from eastern China (12). In 2015, three clinical A. fumigatus isolates harboring TR₃₄/L98H/S297T/F495I or TR₃₄/L98H mutations were identified in Fujian, Nanjing, and Shanghai, respectively (13). All three cities are located in the east or southeast of China. AT present, the epidemiology of azole resistance in A. fumigatus in different parts of China and its association with azole fungicides in agriculture are still largely unknown since susceptibility testing is not routinely performed in clinical microbiology laboratories and azole-resistant strains have not yet been identified from environmental sources (14). In the present study, we sought to investigate the occurrence and characteristics of azole resistance in clinical and environmental A. fumigatus isolates from different geographic areas in China and explore the genetic relatedness of azole-resistant isolates from China and international collections through putative cell surface protein (CSP) and microsatellite genotyping.

MATERIALS AND METHODS

Collection of clinical isolates.

A surveillance of clinical A. fumigatus was conducted in 11 hospitals located in Beijing, Shenyang, Shijiazhuang, Ji'nan, Changsha, Fuzhou, Shanghai, and Chengdu from 2010 to 2015. The geographical distribution of these hospitals is provided in the supplemental file (see Fig. S1 in the supplemental material). A total of 317 A. fumigatus isolates (primarily from respiratory specimens) fulfilling the morphological identification criteria were submitted to the laboratory of the Institute of Disease Control and Prevention, Academy of Military Medical Sciences, for further analysis.

Environmental sampling and isolation of A. fumigatus.

A total of 392 soil samples and leaves were collected from the hospital gardens (n = 162), city parks (n = 196), and farmland (n = 34) located in 13 cities in China between 2014 and 2015 (see Fig. S1 in the supplemental material). The samples were processed as described previously (15). A total of 144 A. fumigatus isolates were cultured and identified according to microscopic and macroscopic morphologies.

Antifungal susceptibility testing and cyp51A gene sequencing.

All 461 A. fumigatus isolates were screened for azole resistance by assessing growth in RPMI 1640 plus 2% glucose agar plates containing 4 mg/liter itraconazole (ITC) and 1 mg/liter VRC according to a protocol described previously for the screening of ITC resistance (16). A subset of A. fumigatus isolates which could not grow on any of the azole-containing agars were randomly selected and subjected to in vitro susceptibility testing and the molecular assay, together with all of the isolates growing on ITR- or VRC-containing agar plates. In vitro susceptibility testing of A. fumigatus to ITC, VRC, and posaconazole (POS) was conducted according to the EUCAST broth microdilution E.DEF 9.3 reference method (17). Azole-resistant or non-wide-type isolates were defined according to the EUCAST E.DEF 9.3 definition criteria. All of the azole-resistant and susceptible control isolates were further identified by growth at 48°C and sequencing of the β-tubulin gene as previously described (18). As a result, all of these isolates were identified as A. fumigatus sensu stricto. The promoter and full coding sequence of the cyp51A gene in these isolates were amplified by PCR, and both strands were sequenced as described previously (19, 20).

Determination of CSP type, microsatellite type, and mating type.

CSP types were determined for all the azole-resistant and susceptible control isolates by PCR amplification and subsequent sequencing (21–23). Nine microsatellite (or short tandem repeat [STR]) loci (STRAf 2A, 2B, 2C, 3A, 3B, 3C, 4A, 4B, and 4C) were also amplified, the fragment sizes were determined, and repeat numbers were assigned as previously described (24). For phylogenetic analysis, microsatellite types of cyp51A non-wild-type A. fumigatus from previous studies (8, 10, 25–35) were included, along with the azole-resistant isolates in this study. The mating types of these isolates were identified by PCR amplification of mating-type-specific genes using two pairs of primers (36).

Ethics statement.

The institutional ethics committees of the Academy of Military Medical Sciences of the Chinese People's Liberation Army, Beijing, China, approved the study. Since all data were anonymously collected and interpreted, the institutional ethics committees waived the need for written informed consent from the participants.

RESULTS

Prevalence and characterization of azole resistance.

In all, 10 A. fumigatus isolates showed the ability to grow on both ITC- and VRC-containing agars (see Fig. S2 in the supplemental material). These isolates were further identified as azole resistant after in vitro susceptibility testing and the detection of cyp51A mutations, including 8 clinical isolates and 2 environmental isolates. The prevalences of azole resistance in clinical and environmental A. fumigatus isolates were 2.5% (8/317) and 1.4% (2/144), respectively. These isolates originated from five different cities, including Fuzhou, Shanghai, Chengdu, Beijing, and Shenyang. All 10 azole-resistant isolates had a POS MIC greater than the breakpoint (≥0.5 mg/liter). Two clinical isolates and two environmental isolates harboring the TR₃₄/L98H/S297T/F495I mutation, as well as five clinical isolates harboring the TR₃₄/L98H mutation, exhibited high-level resistance to ITC (≥16 mg/liter) (Table 1).

TABLE 1.

Characterization of azole-resistant A. fumigatus isolates from China

Strain	Geographical origin	Specimen typea	Mating type	CSP type	STR type	cyp51A mutation	MIC (mg/liter)
							ITZ	VRC	POS
C94	Shanghai	Clinical	MAT1-1	t02	13-10-9-38-9-10-8-10-8	TR34/L98H	≥16	2	1
C96	Shanghai	Clinical	MAT1-2	t04A	23-12-15-23-12-6-13-10-7	TR34/L98H/S297T/F495I	≥16	1	0.5
C116	Fuzhou	Clinical	MAT1-1	t02	13-10-9-38-9-10-8-10-8	TR34/L98H	≥16	4	0.5
C135	Fuzhou	Clinical	MAT1-1	t02	13-10-9-38-9-10-8-10-8	TR34/L98H	≥16	2	0.5
C136	Fuzhou	Clinical	MAT1-1	t02	13-10-9-38-9-10-8-10-8	TR34/L98H	≥16	2	0.5
C195	Beijing	Clinical	MAT1-1	t01	26-21-12-25-9-19-14-9-9	TR46/Y121F/T289A	1	≥16	0.5
C485	Shenyang	Clinical	MAT1-2	t04A	23-12-15-23-22-6-13-10-7	TR34/L98H/S297T/F495I	≥16	2	1
E739	Beijing	Environmental	MAT1-2	t04A	23-12-15-23-22-6-13-10-7	TR34/L98H/S297T/F495I	≥16	2	0.5
C821	Chengdu	Clinical	MAT1-1	t02	18-24-14-11-19-29-16-9-20	TR34/L98H	≥16	4	1
E1001	Fuzhou	Environmental	MAT1-2	t04A	23-12-15-30-24-26-10-10-7	TR34/L98H/S297T/F495I	≥16	1	0.5

^aThere were seven clinical respiratory samples, one wound secretion sample, and two soil samples.

cyp51A mutations in azole-susceptible A. fumigatus.

Among the 153 randomly selected azole-susceptible A. fumigatus isolates, 36 isolates (23.5%) harbored at least one loci of amino acid substitution in cyp51A (Table 2). The most common cyp51A mutation was N248K in both clinical and environmental isolates. The A9T, M172V/A330P, and N248K/M499K mutations were only detected in environmental isolates, whereas the D343N, I242V, K314T, D262Y, F46Y/M172V/E427K mutations were only detected in clinical isolates. Interestingly, the F46Y/M172V/N248T/D255E/E427K mutation, which has been reported to be associated with azole resistance (37), was detected in one clinical isolate from Shanghai and in one environmental isolate from Xinjiang, respectively.

TABLE 2.

Amino acid substitutions in cyp51A, mating types, and in vitro antifungal susceptibility testing results for 153 azole-susceptible A. fumigatus from China

cyp51A mutation	Frequency (no. of isolates)	%	No. (%)		MIC range (mg/liter)
			Clinical isolates	MAT1-1 type isolates	ITZ	VRC	POS
N248K	20	13.1	11 (55.0)	6 (30.0)	0.125–1	0.125–1	0.0625–0.25
A9T	4	2.6	0 (0.0)	4 (100.0)	0.5–1	0.5–1	0.125
D343N	2	1.3	2 (100.0)	1 (50.0)	0.5–1	0.25–0.5	0.25
D257G	2	1.3	1 (50.0)	0 (0.0)	0.5	0.25–0.5	0.125
F46Y/M172V/N248T/D255E/E427K	2	1.3	1 (50.0)	1 (50.0)	0.5–1	1	0.25
M172V/A330P	1	0.7	0 (0.0)	0 (0.0)	0.5	0.5	0.25
I242V	1	0.7	1 (100.0)	0 (0.0)	1	0.25	0.25
K314T	1	0.7	1 (100.0)	1 (100.0)	0.5	0.5	0.125
D262Y	1	0.7	1 (100.0)	0 (0.0)	0.25	0.5	0.0625
F46Y/M172V/E427K	1	0.7	1 (100.0)	1 (100.0)	0.5	0.5	0.25
N248K/M499K	1	0.7	0 (0.0)	1 (100.0)	0.5	1	0.25
None	117	76.5	73 (62.4)	72 (61.5)	0.125–1	0.25–1	0.0625–0.25

CSP typing.

The CSP typing results showed that the five TR₃₄/L98H isolates and four TR₃₄/L98H/S297T/F495I isolates corresponded to t02 and t04A, respectively, whereas the TR₄₆/Y121F/T289A isolates corresponded to t01 (Table 1). There were 15 CSP types in 92 clinical azole-susceptible A. fumigatus and 7 CSP types in 61 environmental azole-susceptible A. fumigatus, with t01, t04A, and t03 being the three most common CSP types in both clinical and environmental isolates (Table 3). One CSP type first identified here was defined as t25. Updated information about the CSP typing nomenclature is provided in Tables S1 and S2 in the supplemental material.

TABLE 3.

Distribution of CSP types observed among clinical and environmental isolates of A. fumigatus in this study and in three previous studies

CSP typea	This study			Chinab	Netherlandsc		Australiad
	Azole resistant (n = 10)	Clinical susceptible (n = 92)	Environmental susceptible (n = 61)	Clinical (n = 162)	Azole resistant (n = 62)	Azole susceptible (n = 55)	Clinical (n = 95)	Environmental (n = 27)
t01	1	34	23	19	4	15	21	5
t02	5	4	7	16	13	4	7	2
t03		12	9	33	1	7	25	4
t04A	4	23	15	51	1	16	24	11
t04B					20
t05		1		2		2	2
t06A		3		5			3
t06B		1		9			1
t07
t08						5
t09						1	1	2
t10		3	4	6		1	2
t11					23
t12		1		3
t13						2		2
t14		1		3		1	1
t15
t16
t17		1		1			1
t18A		3	1			1	2
t18B		2	2	3
t19							4	1
t20							1
t21				1
t22		2		8
t23				1
t24				1
t25b		1

^aCSP type t18A was previously designated as t18. t25 is a new CSP type identified in the present study.

^bGao et al. (22).

^cCamps et al. (40).

^dKidd et al. (21).

Microsatellite typing and phylogenetic analysis.

The 10 azole-resistant A. fumigatus isolates were divided into five different STR types. Three clinical TR₃₄/L98H isolates from Fuzhou and one clinical TR₃₄/L98H isolate from Shanghai shared the same STR type, while one environmental TR₃₄/L98H/S297T/F495I isolate from Beijing and one clinical TR₃₄/L98H/S297T/F495I isolate from Shenyang shared the same STR type (Table 1). Phylogenetic analysis of the 10 azole-resistant isolates and 153 azole-susceptible isolates showed that eight azole-resistant isolates belonged to one major cluster, which consisted of 49.1% of all the surveyed isolates. The clinical TR₃₄/L98H isolate from Chengdu (C821) represented one separate clade far away from other strains, while the TR₄₆/Y121F/T289A belonged to another clade, which mainly consisted of azole-susceptible environmental isolates (Fig. 1).

FIG 1.

Genotypic relationships of 163 representative clinical and environmental Aspergillus fumigatus isolates from China. The dendrogram is based on a categorical analysis of nine microsatellite markers in combination with UPGMA (unweighted pair-group method with arithmetic averages) clustering.

To further explore the genetic relationship of azole-resistant A. fumigatus isolates from different geographical origins, microsatellite typing results of 102 isolates from previous studies (8, 10, 25–35) were included for comparisons. All of the TR₃₄/L98H/S297T/F495I isolates from the present study were closely related to the isolates harboring the same cyp51A mutation from Hangzhou and Taiwan in China (12, 30). The predominant TR₃₄/L98H clone observed here was closely related to the clinical or environmental isolates from the Netherlands and Denmark. This clone belongs to one of the largest clusters among all the non-wild-type A. fumigatus. Some clinical and environmental TR₃₄/L98H isolates from India, Kuwait, Australia, and Tanzania harbored nearly identical STR types, which belonged to another major cluster of the non-wild-type A. fumigatus in the world (Fig. 2; see also Fig. S3 in the supplemental material).

FIG 2.

Genotypic relationship in representative azole-resistant Aspergillus fumigatus from this study and 102 other isolates from the reference according to UPGMA clustering of microsatellite typing results.

Distribution of mating types.

The mating type for all four TR₃₄/L98H/S297T/F495I A. fumigatus samples in this study was MAT1-2, whereas for the other six resistant isolates it was MAT1-1. There were 87(56.9%) MAT1-1 and 66(43.1%) MAT1-2 isolates among 153 azole-susceptible A. fumigatus isolates. The proportions of MAT1-1 mating type among clinical and environmental A. fumigatus isolates were 60.9% (56/92) and 50.8% (31/61), respectively; however, the difference was not statistically significant (χ² = 1.51, P = 0.219). The relationships between the distribution of CSP types and mating types are shown in Table S3 in the supplemental material; these findings suggest that most of the CSP type t02 and t03 isolates correspond to MAT1-1.

DISCUSSION

The emergence of azole resistance in clinical and environmental A. fumigatus has become a worldwide problem in recent years (38). However, large gaps exist in the study of azole resistance in A. fumigatus in China. We present here new epidemiological information, including the first isolation of environmental azole-resistant A. fumigatus from China and the first isolation of clinical azole-resistant isolates from Northeast (Shenyang) and Southwest (Chengdu) China (see Fig. S1 in the supplemental material). Although the prevalence of azole resistance in China was still low compared to the Netherlands (39) and India (30), inadequate diagnosis of aspergillus disease and the lack of susceptibility testing in Chinese hospitals warrant that more effort should be made to narrow this gap.

Our study indicated that TR₃₄/L98H and TR₃₄/L98H/S297T/F495I were the most commonly identified resistance mechanisms in China. In addition, we previously reported the emergence of a new environmental mechanism (TR₄₆/Y121F/T289A) in A. fumigatus from a patient admitted to a hospital in Beijing; no azole was used before the isolation of this strain (10). Unfortunately, we could not provide detailed information regarding the use of azole drugs in the patients from whom the TR₃₄/L98H and TR₃₄/L98H/S297T/F495I A. fumigatus strains were isolated. However, the isolation of two TR₃₄/L98H strains from soil samples of the hospital garden indicated that this mechanism might have originated from environmental routes.

High polymorphisms existed in the cyp51A gene that produced amino acid changes among azole-susceptible A. fumigatus isolates, with N248K being the most common mutation. This mutation was identified in isolates from 10 different cities of China, and 90.0% of these isolates correspond to CSP type t01, suggesting that N248K isolates have a wide geographical distribution and a close genetic relationship in China. It is interesting that many Australian isolates with elevated MICs to azoles displayed the N248T mutation (35). Comparison of CSP type distribution from China and other countries showed that type t04B and t11, which were commonly identified in TR₃₄/L98H isolates from the Netherlands and other European countries (40), had not been found in isolates from China.

Although there were some pitfalls associated with sizing of the obtained PCR fragments, STR typing was commonly used to investigate the genetic relationships of A. fumigatus due to its high discriminatory power and interlaboratory reproducibility (41, 42). Our study showed that clinical and environmental A. fumigatus could not be separated through STR typing since they were widely distributed in different clusters (Fig. 1). Phylogenetic analysis supported the hypotheses regarding clonal expansion of the TR34/L98H isolates in different countries (Fig. 2; see also Fig. S3 in the supplemental material). One TR₃₄/L98H isolate (C821) from Chengdu displayed a distinct genetic background from the other azole-resistant isolates and therefore might have originated from a different evolutionary trajectory. The current data also suggested that TR₃₄/L98H/S297T/F495I isolates were more commonly found in A. fumigatus from China (including Taiwan) and harbored distinct genetic backgrounds from the isolates with the same mutation in the Netherlands and Denmark (Fig. 2).

Increasing evidence indicated that A. fumigatus should be able to generate resistance to antifungal drugs via the sexual cycle to allow evolution in response to environmental change (40, 43). However, a linkage disequilibrium analysis study from the Netherlands indicated that all the azole-resistant A. fumigatus isolates nested within a single, predominantly asexual, population (44). In our study, TR₃₄/L98H and TR₃₄/L98H/S297T/F495I A. fumigatus isolates correspond to different mating types and exhibited distinct genetic background. Therefore, asexual production of exiting azole-resistant A. fumigatus clones might play a major role in the increasing report of azole resistance worldwide.

There were two main disadvantages in our study. First, the clinical diagnosis and outcome of patients with azole-resistant A. fumigatus could not be evaluated due to a lack of relevant information. Second, it was still difficult to analyze the impact of agricultural fungicides use on the emergence of azole-resistant A. fumigatus based on current data from China. However, limited data showed that agricultural fungicide use accounted for around 23% of the total pesticide use in 2010 in China and has been increasing in recent years (45). Being one of the most important fungicides, azole fungicide use in recent years in China is nearly 30,000,000 kg per year (45), which is much higher amount than in the Netherlands, the United Kingdom, and other European countries (46). It should have caused a high selective pressure for the evolution of A. fumigatus in the agricultural environment.

In conclusion, our study provides detailed information on the occurrence and characterizations of azole resistance in clinical and environmental A. fumigatus from China and the genetic relationship of azole-resistant isolates worldwide. These insights are important in designing strategies to combat increasing levels of azole resistance in A. fumigatus.